About Me

Hey! I’m Festus Kahunla, a Machine Learning Engineer with an electrical‑engineering backbone and an obsession for turning language into insight. My work lives at the intersection of Natural Language Processing and real‑world systems — from voice‑driven interfaces to low‑resource chatbots for emerging markets.

Over the last six years I’ve shipped solutions across Africa and the U.S., blending signal‑processing intuition with transformer models, retrieval‑augmented generation, and scalable MLOps. Highlights include a pneumonia‑detection vision system, an African‑slang sentiment analyzer, and a private search engine that never leaves the browser.

Web craftsmanship: I build production‑grade web apps with Next.js + TypeScript on the frontend, containerized with Docker and shipped via GitHub Actions. You’ll often catch me squeezing milliseconds out of a bundle, crafting accessible UI with Tailwind & shadcn/ui, or wiring GraphQL queries straight into React Server Components.

Research focus: large language models, low‑resource NLP, and multimodal language agents. I’m exploring how lightweight adapters and domain‑specific pre‑training can unlock conversational AI for under‑represented languages.

When I’m not wrangling vectors or pixels, you’ll find me reading research papers, open‑sourcing tiny tools, or writing blogs.